Drinking water represents a potential source of lead exposure. The purpose of the present study was to estimate the magnitude of winter-to-summer changes in household water lead levels (WLLs), and to predict the impact of these variations on BLLs in young children. A study was conducted from September, 2009 to March, 2010 in 305 homes, with a follow-up survey carried out from June to September 2011 in a subsample of 100 homes randomly selected. The first 1-L sample was drawn after 5 min of flushing, followed by a further 4 consecutive 1-L samples after 30 min of stagnation. Non-linear regression and general linear mixed models were used for modelling seasonal effects on WLL. The batchrun mode of Integrated Exposure Uptake Biokinetic (IEUBK) model was used to predict the impact of changes in WLL on children's blood lead levels (BLLs). The magnitude of winter-to-summer changes in average concentrations of lead corresponded to 6.55 μg/L in homes served by lead service lines (LSL+ homes) and merely 0.30 μg/L in homes without lead service lines. For stagnant samples, the value reached 10.55 μg/L in 'LSL+ homes' and remained very low (0.36 μg/L) in 'LSL- homes'. The change in the probability of BLLs ≥5 μg/dL due to winter-to-summer changes in WLL was increased from <5% (in winter) to about 20% (in summer) in children aged 0.5-2 years. The likelihood of having BLLs ≥5 μg/dL in young children during warm months was reduced by at least 40% by flushing tap-water.

Objective: To assess the geometric mean of blood lead levels in children aged below 6 years living in sub‐Saharan Africa.Methods: Search strategies were developed in MEDLINE, Web of Science and Google scholar system, supplemented by screening of references from included papers and review articles. Studies including children younger than 6 years residing in sub‐Saharan Africa (SSA) were eligible. The studies were assessed for quality of methods for blood collection, for participants’ recruitment and for quality control for lead analyses. Data were extracted and synthetised by estimating the mean weighted by sample size.Results: A total of 11 148 published papers were identified with only 16 studies conducted in SSA. Only nine of them were relevant of inclusion criteria. The weighted mean of blood lead levels was 13.1 μg/dl. Five of included studies were classified as being less likely to present selection bias, having performed venipuncture and adequate quality control for lead analyses. The weighted mean from analyses involving only these studies was 16.2 μg/dl. Six of the included studies reported the prevalence of blood lead levels ≥10 μg/dl; levels ranged from 7.0% to 70.9%. Only one study reported the prevalence of blood lead levels ≥5 μg/dl (50%).Conclusions: The geometric mean of blood lead levels in sub‐Saharan African children is still elevated relative to the US children. The new standard from the Centers for Disease Control and Prevention (5 μg/dl) implies that important efforts should be directed towards the initiation (or reinforcement) of prevention programmes and thorough research studies in sub‐Saharan Africa.